Adaptive Noise Suppression via Speech Distortion Estimation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current noise suppression systems in audio processing often introduce speech degradation due to fixed noise suppression levels and reliance on signal-to-noise ratios, which fail to accurately predict speech distortion in varying audio environments.
Innovation Solution
An adaptive intelligent noise suppression system that separates audio signals into frequency bands, computes energy estimates, and uses inter-microphone level differences to estimate speech loss distortion, dynamically adjusting enhancement filters to minimize speech degradation while maximizing noise suppression.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Object-affected harmful factors
If fixed noise suppression is applied, then noise is reduced, but speech distortion increases
Solution Approach 1:
The patent applies dynamics by transitioning from fixed noise suppression to dynamic noise suppression that adapts to changing audio conditions. The system continuously monitors the audio signal and adjusts suppression levels in real-time, allowing the noise suppression to vary based on the current speech and noise characteristics, thereby preventing speech distortion while maintaining effective noise reduction.
Solution Approach 2:
The patent implements feedback mechanisms where the system monitors the output of noise suppression and adjusts based on the resulting speech quality. By analyzing the suppressed signal and comparing it against original speech patterns, the system modulates the suppression strength to maintain optimal speech intelligibility while reducing noise, thus resolving the contradiction between noise reduction and speech preservation.
2Object-affected harmful factors
If higher noise suppression is applied, then noise is reduced further, but speech degradation increases
Solution Approach 1:
The patent applies parameter changes by dynamically adjusting the noise suppression parameter (suppression level) based on the audio environment. Instead of using a fixed high suppression level, the system modifies the suppression parameter in real-time according to the detected speech and noise characteristics, allowing optimal noise reduction while preventing speech degradation through adaptive parameter modulation.
Solution Approach 2:
The system transitions from static to dynamic noise suppression by continuously adapting the suppression level to changing audio conditions. The dynamic adjustment mechanism allows the system to provide higher suppression when noise is prominent and reduce suppression when speech is detected, thereby maintaining speech quality while achieving effective noise reduction.
3Adaptability or versatility
If SNR-based suppression is used, then noise suppression is dynamic, but speech distortion prediction accuracy decreases
Solution Approach 1:
The patent applies segmentation by dividing the audio signal into distinct components (speech and noise) and analyzing them separately. By segmenting the signal and examining the statistical properties of each component independently, the system achieves more accurate prediction of speech distortion while maintaining dynamic suppression capabilities. This segmentation allows for more precise control than simple SNR-based approaches.
Solution Approach 2:
The patent replaces the simple SNR-based mechanical approach with a more sophisticated statistical analysis system. Instead of relying solely on the ratio of signal to noise power, the system uses statistical models to analyze the temporal and spectral characteristics of both speech and noise, providing more accurate distortion prediction while maintaining adaptability through statistical parameter estimation.
Data Source
AI summary
Systems and methods for adaptive intelligent noise suppression are provided. In exemplary embodiments, a primary acoustic signal is received. A speech distortion estimate is then determined based on the primary acoustic signal. The speech distortion estimate is used to derive control signals which adjust an enhancement filter. The enhancement filter is used to generate a plurality of gain masks, which may be applied to the primary acoustic signal to generate a noise suppressed signal.